Resource The Official SFA Match Simulation Process + General Sim Queries Thread

[ClarkeM]

No. I'm asking how such a series of results is possible.

How are they not possible?

 

[Tonga Bob]

Qooty is not a 50/50 game, it's a game made up of hundreds of 50/50 moments that result in something at the end. It's complete dumb luck and it must suck but I'm sure the Warriors have positive records against other teams? Perhaps not as drastic as this of course.

This is though, very ****ing funny. 2 wins in the last 23 games has got to be some kind of record surely? I say this earnestly.

I'm not sure what answer you're looking for here?

It's pretty much a bell curve for other teams with normal distribution between 29-69% for wins. The two huge outliers from a Warriors perspective are the Dragons and GF/Premiership representation.

 

[ClarkeM]

It's pretty much a bell curve for other teams with normal distribution between 29-69% for wins. The two huge outliers from a Warriors perspective are the Dragons and GF/Premiership representation.

So it’s possible?

 

[Tonga Bob]

How are they not possible?

They're not impossible, just extremely improbable.

 

[serial_thrilla]

The chance of success of having 9 wins from 37 games (overall CIW record) is 0.0905%, and the chance of 1 win in 15 (my record) is 0.0458%. That's mighty close to zero.

So you're saying that it's possible? Good to know.

 

[Tonga Bob]

Yes it's possible. My question was how is it possible?

 

[Mobbs]

Anyone signs Tonga Bob.

Seriously, it's as you expect. Any one set of results has the same likelihood as any other.

So WWLWLWWLLLWL has exact equal likelihood as LLLLLLLLLL

But i agree it's a nutso ratio. Maybe if you changed name to Tonga Obo, you'd win almost all your matches.

 

[Dingster]

Tiger Bob Tonga Bob there is a form component in the sim for each player and therefore each team based on that simmers' sim. For example the form of my simmed games will be different to that of serial_thrilla and MWPP. That could partially explain a run of form during a season. Hence you often see a team that gets on a roll late in the season goes all the way. A team that starts well (Wazzas of the past come to mind) start well but lose form. Even individually it can happen. I started this season on fire at CHF and was BSing Carey comparisons. Then I got interchanged one game and it turned against me. The sim can be a fickle mistress!

But as Clarkey and Mobbs said a run like that is unusual but not impossible. But one thing to note. Each season the simmers purge their sim and have to re-create the league and each team for the new season. Everything starts fresh.

 

[Dingster]

Also for those wondering or new to Sweet, the admin decides which games the simmers sim and we can not sim our own games.

 

[Jabba73]

Hi Mobbs and the simmer community, long time listener first time melter.

I have a question.

Given the current process for simulation which I agree is run honestly and above board, how do you account for extended cases of one result occuring almost to the exclusion of other results? I have two examples to bring to the table:

The first is the record between CIW & DRA. Overall the result is this:

Wins	9​
Draws	2​
Losses	26​
Total	37​
Win %	24.32%​
Margin	-490​	-13.24 per game​

The first 14 games seems random enough - 7 wins, 1 draw and 6 losses which is a 50% win record. But from season 24 onwards it's 2 wins, 1 draw and 20 losses at 8.7% win rate.

Then since I started in s29 the result is this:

Me
Wins	1​
Draws	1​
Losses	13​
Total	15​
Win %	6.67%​
Margin	-398​	-26.53 per game​

The last 11 results have been zero wins, one draw and ten losses with a total margin of -321 points.

The second example is the fact that my team have played in only one grand final and have won no premierships in 21 seasons (will be 22 this season).

So my question is, how does the current process allow such series of results that have incredibly low probabilities of occurring to happen? Is it dumb luck, or is there something else to it. I do statistical analysis as part of my job, and at a statistical level, this run chart would give me cause for concern

View attachment 2291461

That is a very random set of results.

Nearly as improbable as playing in the rain at both The Underground Stadium and Abdu Prison, yet we did both in one season.

Random

 

[serial_thrilla]

Hate how many players with 175+ games have a worse W/L ratio than Tonga Bob?

 

[serial_thrilla]

Hate how many players with 175+ games have a worse W/L ratio than Tonga Bob?

Holy shit Broken you poor bastard!

 

[Wosh]

That is a very random set of results.

Nearly as improbable as playing in the rain at both The Underground Stadium and Abdu Prison, yet we did both in one season.

Random

What is there? Warm, cloudy, rainy, muddy…
Yes, you will get multiples of those in an 18 game season. The Bears have copped a lot of mud this season, weather doesn’t affect form they are on top of the ladder.

 

[Jabba73]

What is there? Warm, cloudy, rainy, muddy…
Yes, you will get multiples of those in an 18 game season. The Bears have copped a lot of mud this season, weather doesn’t affect form they are on top of the ladder.

But raining underground AND in the desert in one season?

A weather miracle ...

 

[Wosh]

But raining underground AND in the desert in one season?

A weather miracle ...

Blame Mobbs
Always works for me.

 

[Wosh]

Holy shit Broken you poor bastard!

View attachment 2291595

 

[Mobbs]

That is a very random set of results.

Nearly as improbable as playing in the rain at both The Underground Stadium and Abdu Prison, yet we did both in one season.

Random

The prison ground is in the yard, its still exposed to the elements.
But fair point re Underground Stadium. The weather is blocked by 5 floors of shopping complex!

 

[Mobbs]

What is there? Warm, cloudy, rainy, muddy…
Yes, you will get multiples of those in an 18 game season. The Bears have copped a lot of mud this season, weather doesn’t affect form they are on top of the ladder.

Muddy, Rainy, Cloudy, Fine, Warm.

The only difference these variables cause to the Sim is the number of transactions that occur. So Muddy games move slower, Cloudy fastest, over the same period of time, resulting in lower scores and disposal counts.

 

[Wosh]

The prison ground is in the yard, its still exposed to the elements.
But fair point re Underground Stadium. The weather is blocked by 5 floors of shopping complex!

Pretty sure Sweet FA safety guidelines dictate they must have fire sprinklers in there so the sim might get confused. It’s not very smart.

 

[Mobbs]

But raining underground AND in the desert in one season?

A weather miracle ...

Half the players are based in Melbourne. I don't think many of them would bat an eyelid at this variability

 

[Mobbs]

Pretty sure Sweet FA safety guidelines dictate they must have fire sprinklers in there so the sim might get confused. It’s not very smart.

Underground Stadium also gets backwash from the Sweet Sewerage Network.

 

[Wosh]

Underground Stadium also gets backwash from the Sweet Sewerage Network.

Top notch engineering right there.

 

[ClarkeM]

Yes it's possible. My question was how is it possible?

How is what possible though?

 

[Jabba73]

The prison ground is in the yard, its still exposed to the elements.
But fair point re Underground Stadium. The weather is blocked by 5 floors of shopping complex!

Keep asking them to fix their leaking sewerage system.

Not convinced it was actually rain

 

[Braklet]

No. I'm asking how such a series of results is possible.